The conditions for user-owned technology are here.

• There is a global collapse of trust in institutions. People have lost faith in institutions that were once core to economic, political, and social life: governments, banks, media, and schools. This is not a short-term trend or a response to any one event. This is a long-running shift in expectations. People no longer assume institutions are neutral, reliable, or aligned with their interests.
• Distributed systems and cryptography give builders new tools to operate without trust. These technologies are meant to operate in adversarial environments: they assume participants may be malicious, software must be verifiable, and systems should still function when counterparties fail.
• AI makes this shift toward trust-minimized systems more urgent and more possible than ever. AI not only centralizes power, but also lowers the cost to build. Now, one person can build in hours what once took teams months. This puts pressure on middlemen, opens new possibilities for builders, and increases demand for infrastructure that puts users in charge.
• User-owned systems defend freedom. User-owned systems minimize trust by shifting control back to the user. These systems cannot be unilaterally changed. They let people build without asking for permission. Done right, they allow users to opt out of systems that no longer serve them without loss in functionality.

Electric Capital invests $1M to $20M in user-owned technology that gives people control, privacy, and access.

Since 2018, we have invested in systems that reduce reliance on intermediaries. We started with programmable money. Today, the same principles and technology apply across software, data, markets, and more.

If you’re building in these areas, we want to invest in your ideas. For deeper background on our thesis, read our 2018 papers on Reimagining Trusted Intermediaries and Programmable Money.

This article outlines 26 opportunities in 2026 across key areas.

These opportunities span user-owned systems, globally accessible markets, entertainment built on new financial primitives, and infrastructure for a world where AI builds software. But they all share a common thread: they explore how power, access, and ownership should work in a world where AI is ubiquitous and deeply embedded.

The opportunities are in six key areas:

• Personal Software: AI makes it possible to build tools tailored to your life, not just adapt to SaaS built for the average user. Private agents, encrypted collaboration, and locally-run software are now feasible and increasingly necessary.
• Agent-Focused Infrastructure: As AI agents become the primary builders of software, the development stack breaks. New primitives are needed for testing, deployment, payments, data access, and coordination between agents.
• Fintech & DeFi: Stablecoins give 4B+ people access to the dollar. Now, they want yield, equity exposure, insurance, and more. Demand for financial infrastructure that is global, programmable, and accessible is accelerating.
• Finance as Entertainment: Younger generations treat markets like entertainment. Trading is fast, social, and fun. This changes the shape of financial products and opens the door to new kinds of markets.
• Metaverse Revival: World models and generative AI collapse the cost of building immersive, personalized environments. People will step into experiences shaped around them rather than passively consume content. There are opportunities in building platforms that simplify world creation and give users control over how their data in these worlds are shared, stored, and monetized.
• New Crypto Primitives and Applications: Proof-of-stake and proof-of-work are maturing and leave room for new consensus models. Zero-knowledge systems and fully homomorphic encryption are becoming practical. These primitives unlock new design space: consensus tied to human or physical input, privacy-by-default infrastructure, and applications built on top for regulated entities, energy markets, and even new jurisdictions.

If you are building in one of these areas, get in touch with us through info@ electriccapital.com.

Personal Software

For the first time, individuals can build software tailored to their exact needs rather than be limited by what corporations offer. Because AI agents can now handle complex workflows like reading emails, scheduling meetings, and managing files, there is new demand for data privacy, data ownership, and data durability. Crypto-enabled systems can make these tools private, durable, and multiplayer.

Specific ideas we want to invest in:

1. Private AI agents: People need to run AI on sensitive data in a secure way.

What it could look like: An AI assistant that automates your personal workflows while preserving your privacy. Connect your health and finance records and get AI insights. The AI models run in a trusted execution environment (TEE) or on a compute network where inbound queries are anonymized. Responses come back without a corporate provider or malicious actors seeing your data.

2. Encrypted collaboration spaces: People need to collaborate privately with others: both humans and agents. Remember the “cloud” is just someone else’s computer.

What it could look like: A shared workspace for friends, families, or small businesses. Finances, documents, and tasks sync via peer-to-peer storage solutions. Selective disclosure authorizes agents to access specific data types. There is no account creation, no corporation to read, store, or train on sensitive data, and it works offline.

3. Desktop agent: People need automation for data on their local computers.

What it could look like: An agent on your desktop runs locally to read your emails, write responses, create agendas, and organize your life. This idea could extend to a new kind of operating system for your desktop in an AI-first world.

4. Pay for services privately: People need ways to pay for software services without identifying themselves.

What it could look like: Buy a VPN, games, cloud storage, or AI compute without an account. You pay per use, metered by the service, and payment is settled in stablecoins via x402 or similar protocols. The service knows someone paid and how much, but not their identity.

Agent-Focused Infrastructure

Agents will write most of our code, and perform most of our knowledge work. Key implications: (1) Software tools need to be rebuilt from the ground up because AI-generated code introduces new failure modes. (2) Development will move in-house because custom software now makes economic sense. (3) Agents need new rails to transact with each other. (4) Businesses once limited by human labor can suddenly scale. These ideas capture opportunities from these second order effects.

Specific ideas we want to invest in:

• AI-native compute infrastructure: Companies need the ability to test, isolate, and revert AI-generated changes at the infrastructure level.

What it could look like: AWS or GCP rebuilt for agents. Agents write code in sandboxes, test against production data safely, and deploy with automatic rollback if something breaks. The entire flow assumes code comes from agents and not humans.

• End-to-end product development tools: Non-technical employees need to go from idea to working software.

What it could look like: A platform where a user specifies a business goal, data sources, and desired outcomes. The system produces plans, designs, code, and a working product. The system removes the need for technical translation, allowing non-technical users to move directly from “idea” to “deployed product” in hours rather than months.

• Agent-enabled commerce: Agents need to buy and sell autonomously without human identity or bank accounts.

What it could look like: An API marketplace where agents buy services from other agents. Discovery, negotiation, and payment per call use protocols like x402 to settle instantly with stablecoins.

• Data networks and marketplaces: AI needs data infrastructure that compensates contributors and gives them control over usage.

What it could look like: A network where users share medical records, spending patterns, investment behavior, or creative works for AI training. Contributors set permissions and get paid when their data improves models. AI companies get the financial data they need with clear provenance.

• Scaling professional services: Service businesses need AI-native operations to scale beyond what human labor allows.

What it could look like: A law firm where every attorney has an AI associate handling research, drafting, and document review. A firm that served 1,000 clients now serves 100,000. Any client services profession—lawyers, architects, marketers, accountants, financial advisors—can be rebuilt with AI at its core.

Fintech & DeFi

4B+ people and millions of businesses who face currency risk are actively seeking dollar access through stablecoins, representing the largest expansion of the dollar’s network effects in decades. As stablecoins provide access to dollars to individuals around the world—growing from $3B in 2019 to over $300B today—millions of new dollar holders need more than just digital cash. They need yield, investment opportunities, and financial services. There are growing opportunities in financial products that give users ownership and global access.

Specific ideas we want to invest in:

• Non-crypto correlated yield: Stablecoin holders need yield that does not drop when BTC price drops.

What it could look like: A platform that brings real-world infrastructure yields to stablecoin holders. Yield could come from data center project bonds, solar installations, and EV charging networks with predictable cash flows and no correlation to crypto.

• Globally accessible equities: Global investors need direct ownership of foreign opportunities with low friction and costs.

What it could look like: A financial product that replicates equity ownership with price exposure, no funding rate, and no expiration. A trader in the Philippines builds a US tech portfolio. A Canadian builds Korean semiconductor exposure.

• New forms of insurance: Businesses need fast, transparent coverage for operational risks that traditional insurance cannot provide.

What it could look like: A platform that uses prediction markets to create new insurance products. A hotel chain can buy hurricane protection for its Florida properties. A ski resort can hedge against a warm winter. Capital providers supply liquidity in exchange for uncorrelated yield.

• Commodities markets on-chain: Commodities need markets with 24/7 trading, instant settlement, and global access.

What it could look like: A market to trade energy storage capacity. Battery storage capacity is one potential starting point because data centers need reliable power and are investing in storage to reduce grid dependence and integrate renewables. A data center with excess storage capacity can sell storage to a neighboring facility during peak demand. Grid operators can trade capacity based on seasonal needs.

• Protected DeFi assets: Institutions need to deploy assets into DeFi that can be safe despite hacks.

What it could look like: A wrapped version of ETH that can be reverted if a protocol gets hacked (GuardedETH). A trusted committee reviews exploits and can reverse the GuardedETH without the underlying ETH ever moving. Legitimate transactions proceed normally.

Finance as Entertainment

Younger generations see financial markets as a meritocratic alternative to traditional paths. As they participate in markets, they reimagine what markets look like and turn them into entertainment. They trade like they play games: they look for high-adrenaline trades with fast feedback loops in markets that are easy to learn and accessible. Fast-turnaround products like zero-days-to-expiration options, which can settle in hours, now exceed 55% of S&P 500 options volume. Easy-to-access markets like prediction markets, where anyone can put money on news headlines, hit $44B in volume in 2025, up 5x from the prior year. They’re also turning their trades into content: positions are discussed on Discord in real time, wins and losses are shared on TikTok, portfolios are reviewed on Twitch. When markets become entertainment, there are opportunities for new platforms that treat financial data like they their users do: as fun, participatory content.

Specific ideas we want to invest in:

• Spectator capital: Live stream audiences need ways to financially participate in outcomes.

What it could look like: A platform that lets audiences stake on live content. Participation makes watching more fun, but currently viewers are limited to tips and subscriptions. This platform lets reality TV audiences stake predictions on who gets eliminated, or lets audiences copytrade as a streamer shares their trading session.

• Opinion markets: Prediction market platforms need markets that resolve on collective belief, not just events.

What it could look like: A platform that generates market-ranked lists. Users stake positions on where they believe others will rank these items. This platform creates lists for the best pizza in NYC, the top wines under $20, the most influential movies of the last decade, or the best AI developer tools, all ranked by the market. Lists resolve weekly based on stake-weighted rankings.

• Drama launchpads: Individual storytellers need funding and distribution platforms now that they can produce dramas more cheaply than studios.

What it could look like: A UGC platform for short-form drama. Creators produce episodes with AI video tools: mafia boyfriend sagas, secret billionaire reveals, revenge thrillers. Fans unlock episodes with tokens and tip creators directly. Creators earn based on viewership. ReelShort generated $700M+ revenue in Q1’25 with low-budget, studio-made drama series. This platform combines YouTube UGC content with ReelShort’s video format.

Metaverse Revival

Immersive digital worlds are now economically viable to build. Over the past two years, AI models for image, video, and simulation have advanced rapidly, collapsing the cost of creating assets and environments. Solo creators can now build what used to need whole game studios. At the same time, demand for personalized, interactive content is accelerating: Dispatch, a choose-your-own-path TV/game hybrid, sold 3.3M copies in 3 months for $85M with a 98% approval rating. Roblox saw a 70% YoY increase in DAUs and paid out $428M to creators in Q3 2025 alone. Personalized, AI-driven character chat apps like Character AI also show strong early demand for individualized entertainment. These new environments will not just entertain users, they will also generate rich, structured interaction data for world models and robotics.

Specific ideas we want to invest in:

• World compiler: Solo creators without specialized skills need tools that turn natural language into fully interactive 3D environments.

What it could look like: A platform that turns natural language into fully interactive 3D worlds. Building 3D environments still requires specialized skills in modeling, physics, and NPC behavior. AI can collapse this barrier. A creator describes a world and the system builds it. Assets, physics, NPC logic, and memory are handled automatically. A solo creator ships a rich virtual environment in days instead of years.

• Procedural narrative engine: Players need stories that adapt to them and never end.

What it could look like: A platform that generates player-specific stories in real time. Linear stories have endings. Players want experiences that adapt to them and keep going. A user enters a detective universe and every case is unique to them. Characters remember past interactions. Plot twists respond to their choices. The story never runs out.

• World-as-dataset platform: World models and robotics systems need diverse interaction data. Consumer immersive environments generate this data constantly, but no one is capturing it.

What it could look like: A VR game where every player interaction is instrumented. How users navigate rooms, pick up objects, and interact with characters become training data for robotics. Users opt in, set permissions on what data is shared, and get compensated. AI companies get real human behavior data they cannot generate synthetically.

New Crypto Primitives & Applications

Crypto primitives are no longer theoretical. Proof-of-stake and proof-of-work have both proven resilient at scale. Zero-knowledge (ZK) proofs are moving out of research and into production systems. Fully Homomorphic Encryption (FHE) is getting faster and more usable. As these foundational technologies mature, they unlock new opportunities for builders to create systems that prioritize privacy, embed real-world inputs into consensus, and support coordination for legacy systems like energy markets or governments.

Specific ideas we want to invest in:

• Human time as consensus: Blockchain networks need consensus mechanisms anchored to human effort, not just capital.

What it could look like: Proof of useful work, where consensus requires completing tasks with external value like labeling data or verifying real-world events. Participation rights flow from demonstrated capability rather than stake.

• Physical resource networks: Small-scale infrastructure operators need coordination systems that make their contributions economically rational.

What it could look like: Energy networks where production or storage is consensus weight, aligning grid stability with network security. Sensor networks anchored to physical measurement like weather, water, or infrastructure monitoring.

• Privacy-native L1s: Healthcare, enterprise, regulated finance, and many other industries need chains where privacy is the default.

What it could look like: Confidential state machines where computation happens on encrypted data by default. Current chains are transparent by default, but many entities like healthcare, enterprise, and regulated finance legally cannot operate on transparent chains. Validators verify without seeing transaction contents using ZK-native architectures or FHE-based execution.

• Use-case specific FHE: Institutions often need to collaborate on data without revealing it to each other.

What it could look like: Banks detect suspicious patterns across institutions without sharing customer data. Each bank runs FHE queries against encrypted data from other banks. They identify accounts that touched the same suspicious entity without revealing their customer lists to each other.

• Energy contract settlement: Traditional markets need crypto rails to create 24/7 settlement between different parties. Deregulated energy markets are a good place to start because they are antiquated and increasingly strained as energy demands from AI increase.

What it could look like: A shared settlement layer for energy contracts in deregulated markets. Delivery data triggers automatic payments. Suppliers see cash flow in real-time. Brokers get their cut instantly. No single party controls the ledger.

• Crypto-native jurisdictions: Special economic zones and frontier jurisdictions need new ways of thinking about governance and financial infrastructure.

What it could look like: A new jurisdiction that adopts crypto rails from day one. On-chain identity, programmable courts, tokenized capital markets, and smart contract-based regulatory logic.

About Electric Capital

Electric Capital is an early-stage venture firm with $3B in assets under management with six funds. Founded in 2018, the firm is engineer-led, with deep expertise in cryptography, distributed systems, machine learning, incentive design, marketplace dynamics, and privacy-preserving technology. Our team is over 30% former founders who have started eight companies (six acquired) and shipped products used by billions.

We have invested in 50+ startups with a cumulative market cap of $50B+. At Electric, we are investors in industry-leading companies such as Anchorage, Aven, Bitnomial, Bitwise, EigenLayer, Ellipsis, Kraken, Monad, Re, SF Compute, Solana, and Spruce. Prior to Electric, our partners were early investors in Airtable, Boom Supersonic, Cruise, Figma, Notion, and Pulley.

Disclaimer:

1. This article is reprinted from [electriccapital]. All copyrights belong to the original author [Avichal Garg, Curtis Spencer, Ken Deeter, Maria Shen, Ren]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.